RU2538031C2 - Method for highly directional reception of sound waves - Google Patents

Abstract

FIELD: physics, acoustics.

SUBSTANCE: invention relates to acoustics, particularly to methods for highly directional reception of sound. A method for highly directional reception of sound waves, wherein reception is carried out with four microphones mounted on a rigid linear base. The zero direction of the receiver is defined by the normal position of the axis of the sound source, and outputs of microphone pairs are connected to two inputs of an operational amplifier through electrical filters tuned to frequencies which correspond to the critical band of the human ear. The number of operational amplifiers is not less than three, and the outputs of the microphones with maximum distance in between L_max are connected to the operational amplifier through low-pass filters; microphone pairs located at a distance L_mid and L_min, are respectively connected through mid- and high-pass filters; output signals of said adder operational amplifiers are used to generate a single output signal, and values of said distances are set based on the expression $$L=\frac{\Delta {\lambda }_{sw}}{{С}_{sw}},$$

where Δ_sw is the phase shift on output microphone pairs with deviation of the sound wave front from the zero position by an angle ±α, expressed as a fraction of the sound wavelength Δλ_sw=C_sw/ƒ_sw, C_sw and ƒ_sw are the speed of sound and frequency of the sound signal component, respectively.

EFFECT: high sound quality.

6 dwg

Description

A method for sharply directed reception of sound waves is proposed, which can be used as sound receivers, as well as in a number of specific acoustic applications of a scientific and research nature.

The known method [1, 2] sharply directed sound reception in the design shown in figure 1, as a tube-type microphone and in the form of an interference microphone of the "traveling wave" type in figure 2 in the form of a single tube with holes.

When implementing the known method in the design of FIG. 1, a sound wave 1 with a front 2 acts on the inputs 3 of the sound channels-tubes 4 with a linearly varying length. The opposite ends of the tubes 4 are assembled into a single end plane and placed in the housing 5 of the microphone 7 in its precapsular volume 6. The microphone 7 can be electrodynamic or condenser and receives sound pressure that enters the volume 6 only from the sound tubes 4.

The known method in another embodiment of FIG. 2 is implemented by one sound channel-tube 4 with acoustic sound receivers in the form of holes 8 in the tube, which, as in FIG. 1, are linearly located along the length of channel 4, including the entrance at the end 3 tube. The channel tube is introduced into the housing 5 of the microphone 7 in the precapsular volume 5.

If the front of the sound wave 2 is normal to the axis of the sound tubes, then the sound pressure arrives at the receiving surface of the microphone 7 from the inputs 3 in figure 1 or from the inputs 8 in figure 2 with the same phase and their amplitudes add up.

This position of the microphone is called zero.

Sound waves that enter the sound channels at an angle of ± α to the zero position act on the receiving surface of the microphone for a time delayed Δt _ass , i.e. turn out to be phase shifted. The maximum delay occurs between the extreme sound receivers at α = 90 °.

So at α = 90 °, the value of Δt _ass is equal to Δt _ass = d / Sv, where d is the difference in the length of the channels in Fig. 1 or the distance between the holes in Fig. 2, Sv is the speed of sound.

Thus, when the angle α changes from one extreme position α = 0 to the other extreme position α = 90 °, the delay time Δt _ass changes from zero (at α = 0) to the maximum.

At 0 <α <90 °, the delay value can be written as (see Figs. 1 and 2):

, откуда $$\Delta t_{s\mathrm{but}d}=\frac{d}{C_{s\mathrm{at}}}-\frac{d\cdot \cos \alpha }{C_{s\mathrm{at}}}$$

from where

. $$\Delta t_{s\mathrm{but}d}\cdot C_{s\mathrm{at}}=d\left(\mathrm{one}-\cos \alpha \right)\left(\mathrm{one}\right)$$

.

The product Δt _ass · Szv has a dimension of length and, therefore, in relation to sound vibrations is comparable with their wavelength (λ = Szv / fsv is the frequency of sound vibrations), i.e. 180 ° phase delay, for example, corresponds to a phase delay of 0.5λ sv.

Expression (1) characterizes the known method and allows for given values of α and d to calculate the frequency of sound vibrations fzv at which there will be a full compensation of sound pressure in the microphone on its receiving surface. So, at the lower frequency of the sound signal fnch = 340 Hz, 0.5λvz = 0.5 m, compensation of sound vibrations on the microphone’s receiving surface at α = 20 ° is possible if d between the acoustic inputs is equal to:

, $$d=\frac{0,5{\lambda }_{s\mathrm{at}}}{\mathrm{one}-\cos \alpha }=\frac{0,5}{\mathrm{one}-0,9397}=\frac{0,5}{0,06}=8,3m$$

,

and at α = 45 ° d≈1.7 m.

With increasing fsv, the length d of course decreases, however, for reporter tasks and receiving speech signals, the sound receiving channels in the known method turn out to be either sufficiently long or, when they are shortened, lose their general directivity.

This drawback is the first in the known method of directional sound reception.

Another disadvantage of this method is the appearance in the sound channel (especially in figure 2) of spurious sound vibrations that arise as a result of reflections of incoming sound waves from the receiving surface of the microphones and from the holes in these channels.

It is known [3] that almost all wind musical instruments are built on the emergence of resonant frequencies due to the reflection of sound waves from inhomogeneities in the form of holes in the sound channel, which are equipped with movable valves.

According to its technical characteristics, the known method for the formation of sharply directed sound reception according to [1 and 2] can be considered as a prototype (analog) of the proposed method.

The technical solution to the proposal is to address these shortcomings:

- change the time delay Δt _back used in the method [1, 2], to another dependence, which would provide a decrease in the linear size d with the same value of α;

- an exception when receiving sound waves of possible resonant spurious oscillations.

The proposed method is a sharply directed reception of sound waves, in which sound is received by sound wave receivers located on a rigid linear basis. In accordance with the proposal, at least four microphones are used as sound receivers, the zero direction of the receiver is determined by the normal position of the axis of the sound source - the middle of the mentioned base with microphones, while the microphone outputs are connected in pairs to the inputs of operational summing amplifiers, the output signals of which are used to form a single output signal, and the distance between pairs on a linear basis is determined from the expression:

, $$L=\frac{\Delta {\lambda }_{s\mathrm{at}}}{\sin \alpha }$$

,

moreover, in the microphone - input circuit of the amplifier of each pair of microphones include electric filters for frequencies that correspond to the main critical frequency bands of the human hearing so that in the circuit of a pair of microphones with a maximum distance between them Lmax include filters for the low-frequency component of sound, and in the chain of pairs of microphones with the average and minimum distance between pairs Lcp and Lmin include filters for the medium and high-frequency components of sound, where Δλsv is the phase shift at the inputs of a pair of microphones when the sound wave front is deflected m zero position by an angle of ± α, expressed in fractions of the wavelength of the frequency λzv λzv = SCB / fzv and SCB fzv and accordingly the speed of sound and the sound frequency component.

The essence of the proposed method is illustrated by diagrams and drawings:

figure 1 - sharply directed receiver of sound of a tubular type, where 1 is the sound wave, 2 is the wave front, 3 is the sound wave receivers, 4 is the sound channel tubes, 5 is the precapsular volume, 6 is the microphone protective housing, 7 is the microphone receiving surface (actually the microphone);

figure 2 - sharply directed receiver of sound waves of the type "traveling wave", where 8 - holes - receivers of sound waves, 9 - microphone;

figure 3 - sharply directed receiver of sound waves that implements the proposed method, where 10 is a rigid linear base in the form of a rectangular tube, which is shown in sections DE and MN, 11 is a special hollow handle, 12 are structural elements for mounting on various supports; in the cross section of the base 10 according to DE 13 - a glass with a microphone 9, inserted into the base, 14 - a hole in the glass 13 for the conclusion of the wires;

figure 4 - curves of the selectivity of human hearing (critical frequency bands of hearing from [6]);

figure 5 - probability density of the occurrence of formants (from [6]);

6 is a functional diagram of the connection of the main elements according to the proposed method, where 15-15 are filters in the circuit of a pair of microphones with Lmax, 15'-15 'are filters in the circuit of a pair of microphones with Lcp, 15' '- 15' 'are filters in Microphone pair circuit with Lmin; 16 is a summing operational amplifier.

As follows from figure 3, the sound wave 1 with the front 2 acts on the receiving surface of the microphones 9, which are rigidly installed in the holes 8 of the rigid line 10. The line is equipped with a special handle 11, which is rigidly connected to the line. The handle 11 is hollow inside to accommodate amplifiers, filters and power supplies, and on the outside it is provided with structural elements 12 for mounting it on various external supports.

In figure 3, the handle 11 is located in the conditional middle of the line 10, however it can be located on any part of the line and even outside the microphones 9, for which the length of the line should be increased, however the middle of the linear base between the microphones Lmax is saved as the middle of the receiver. The cross section of ruler 10 for DE and MN is also shown. A cross section DE shows a possible embodiment of a rigid installation of a microphone 9 in a beaker 13 (round shape if an electret condenser in the form of a round disk is used as a microphone). The glass 13 is inserted into the hole 8 of the line 10, providing a rigid fixed position of the glass in the hole. The section MN shows the execution of the line 10, in particular, in the form of a hollow pipe of rectangular cross section. It should be noted that the ruler can be metal or plastic, and in cross section from a circle to a rectangle. The dimensions of the ruler are determined by the external dimensions of the microphones and the recorded frequencies of sound waves fzv.

At the bottom of the glass 13 make holes 14 for wires.

The zero direction in the directionally receiving sound in the proposed method in figure 3 corresponds to the position when the axis of the sound source is the middle of the base with microphones, normal to the linear base itself (that is, the front of the sound wave is parallel to the base line). At α = 0, the sound front acting on the microphones will have a single phase.

With the lateral arrival of sound wave 2 to any pair of microphones 9 shown in Fig. 3, that is, for α ≠ 0, its effect is not in-phase, but with a time delay Δt _ass in one of the pair's microphones. From figure 3 it follows that the maximum value Δt _ass occurs in a pair of microphones with a distance Lmax between them. From the triangle ABC in figure 3 it can be seen that the delay value - the side BC of the triangle ABC - can be written in the form:

, $$BC=L\cdot \sin \alpha \left(2\right)$$

,

or, expressing the AC through the value of the time delay Δt, the _back of the sound wave entering the second microphone of the pair, expression (2) takes the form:

$$\Delta {\lambda }_{s\mathrm{at}}=\Delta t_{s\mathrm{but}d}\cdot C_{s\mathrm{at}}=L\cdot \sin \alpha \left(3\right)$$

The value of Δt _ass · C _sv = Δλ _sv can be considered as the phase shift of the action on any pair of microphones of the sound wave, expressed in fractions of the sound wavelength λsv = Svs / fsv.

We use (3) and calculate the distance L. Using the data for calculating the distances d in the well-known analogue to calculate, namely: α = 20 °, sinα = 0.342, Δλzv = 0.5Svz / fvz at Svz = 340 m / s and fzv = 340 Hz.

. $$L=\frac{0,5\cdot C_{s\mathrm{at}}}{f_{s\mathrm{at}}\cdot \sin \alpha }\cong \mathrm{one},5m$$

.

Thus, the proposed method allows to reduce the linear size of the sound receiver by 5.5 times compared with the analogue when using expression (1).

Any sound receiver that converts sound vibrations in the air into an electrical signal is similar to a hearing aid created by nature in humans. The peculiarity of human hearing is that the perceived complex sound signal is analyzed in frequency, amplitude and with certain temporal characteristics [4]. The sharp focus of receiving sound waves is a specific characteristic. The inefficiency of such receivers when recording, for example, musical sounds, is obvious, but a conversation under the same conditions can be “heard”.

Further, the analysis of the proposed method will be carried out on the example of the reception of human speech.

The speech tract is a complex continuously tunable filter with a set of resonances that are created by the mouth, nose and nasopharynx. The monotonous spectrum of the pulses of the fundamental tone that occurs in the vocal cords is converted into a spectrum with maxima and minima. Such maxima in the spectrum of speech are called formants [2, 4, 5]. In this case, the first formant F ₁ occupies the frequency band 270-700 Hz, the second formant occupies the band 600-2000 Hz, the third formant occupies the band 2300-3000 Hz. Figure 4 presents the curves of the selectivity of human hearing from the work [6], which are called the critical frequency bands of hearing. Figure 5 presents the probability density curves of the occurrence of formants in Russian human speech from [5].

Based on the foregoing, the authors consider it appropriate to use electric filters for the purpose of a microphone-operational summing amplifier, as shown in Fig.6, corresponding to the critical frequency bands of human hearing. Since speech is characterized by three main formants, the number of filters should be three.

For specific, for example, biological problems of recording sounds from any biological objects in noise conditions, the number of filters will be determined by the critical frequency band of the expected sound source and, therefore, the number of microphones used will also be different.

According to the authors, the proposed method can be used for sharply directed reception of sound waves in the frequency regions that a person does not hear - in infra- and ultrasound.

Figure 6 shows that a pair of microphones 9 with a distance Lmax should ensure that the filters 15-15 emit the low-frequency component of sound waves (corresponding to the formant Ф ₁ , in the band 270-700 Hz with an average frequency in the band ~ 500 Hz). Pairs of microphones 9 with distances Lcp and Lmin are equipped with filters that provide amplification of the components of the audio signal in the frequency band of the formant F ₂ (with an average filter frequency of 1300 Hz) and in the frequency band of the formant F ₃ (with an average filter frequency of ~ 2700 Hz).

The distance between the microphones 9 on the basis of 10 is set by setting the necessary (required) distance L. For example, if Lmax for some reason should not exceed 1 m, then the angle α at which the signals in the summing amplifier from the two microphones will be completely compensated by f _cfF1 = 500 Hz, will be 20 °. Is it possible to reduce this angle? It is possible, if, for example, f _{fPF1 is increased} to 600 Hz, then at L≈1 m α is equal to ~ 17 °.

The fixed distance L corresponding to the fcp formants f ₁ , f ₂ or f ₃ creates the conditions under which the extreme frequencies within the formant band are not fully compensated in the summing amplifiers, but partially. From the section of trigonometry it is known that the sum of two sinusoidal quantities with the same frequency ω is also a sinusoidal value with the same frequency:

A ₁ sin (ωt + φ ₁₎ + A ₂ sin (ωt + φ ₂₎ = Asin (ωt + φ) ,

where is the total amplitude $$A=\sqrt{A_{\mathrm{one}}^2+A_2^2+2A_{\mathrm{one}}{\mathrm{<figure-callout\; id="2"\; label="A"\; filenames="00000019.png,00000020.png"\; state="\{\{state\}\}">A</figure-callout>}}_2\cos \left({\varphi }_2-{\varphi }_{\mathrm{one}}\right)},\left(\mathrm{four}\right)$$

.and phase $$\varphi =\frac{A_{\mathrm{one}}\sin {\varphi }_{\mathrm{one}}+A_2\sin {\varphi }_2}{A_{\mathrm{one}}\cos {\varphi }_{\mathrm{one}}+A_2\cos {\varphi }_2}\left(\mathrm{four}\right)$$

.

The case of exposure of a sound front to a pair of microphones 9 shown in FIG. 3 occurs under conditions when A ₁ = A ₂ (a single front), and one of the phases, for example, φ ₁ , can be taken equal to zero (reference point). Then expression (4) can be written as:

$$\begin{array}{l}A=\sqrt{2A_{\mathrm{<figure-callout\; id="2"\; label="one"\; filenames="00000019.png,00000020.png"\; state="\{\{state\}\}">one</figure-callout>}}^2+2A_{\mathrm{one}}^2\cos {\varphi }_2}\left(5\right)\\ \varphi =\mathrm{<figure-callout\; id="2"\; label="t\; g\; \phi "\; filenames="00000019.png,00000020.png"\; state="\{\{state\}\}">t\; </figure-callout>}g{\varphi }_{}{}_2\end{array}$$

When changing the angle φ ₂ from 0 ° to 90 °, cosφ ₂ varies from 1 to 0, when changing φ ₂ from 90 ° to 180 ° cosφ ₂ changes from 0 to -1, when changing φ ₂ from 180 ° to 270 ° cosφ ₂ varies from -1 to 0 and when changing φ ₂ from 270 ° to 360 ° cosφ ₂ changes from 0 to 1. The given changes of φ ₂ show the corresponding changes of A.

Let us calculate as an example a pair of microphones with a distance between them Lmax and with low-pass filters at the frequency of the formant Ф ₁ with f _cr.F = 500 Hz the amplitude of the summing signal A at the amplifier output at the extreme frequencies of the formant Ф ₁ - 270 Hz and 700 Hz.

At the critical frequency λ _zv.kr ≈0,7 m and α = 30 ° (sinα = 0,5) Δt _backside SCB = 0.5 λkr = 0.35 m, and for fzv = 270 Hz λzv = 1, 26 m and for fsv = 700 Hz λsv = 0.48 m and then the phase delay for the frequencies fsv and fsv '', translated into degrees, is 100 ° for fsv 'and 262.5 ° for fsv''. At these angles, the amplitude A from (5) is:

и для A_{700 Гц}≅A₁ $$A_{270Gc}=\sqrt{2A_{\mathrm{one}}^2-2A_{\mathrm{one}}^2\left(\mathrm{one}-0,9848\right)}\cong A_{\mathrm{one}}$$

and for A _{700 Hz} ≅A ₁

Since at α = 0 the amplitudes at the output of the summing amplifier are 2A ₁ , then when the threshold level is introduced at the inputs of the summing amplifiers, for example, at the amplitudes ~ A ₁ , then the extreme frequencies of the speech formants are completely compensated in the angle of the considered example ~ 30 °.

The proposed method also by reducing the frequency band of hearing formants (for example, for the low-frequency formant F ₁ from 350 to 600 Hz) with the introduction of the mentioned threshold level at the inputs of summing amplifiers will allow you to form the direction of sound reception in the angle ± (10 ° ÷ 15 °). This statement by the authors is also based on the fact that the semantic survivability of a speech signal is unique.

So, when processing the clipping of an analog speech signal, intelligibility does not deteriorate. The amplitude envelope of the speech signal, separated from the signal itself, also retains intelligibility. Even if in the instantaneous spectrum of a speech signal there is only one frequency having a maximum amplitude, then intelligibility is quite acceptable [4, 5].

A similar calculation was performed by the authors for other critical hearing bands, and the calculation result was almost the same. The conclusions made about possible combinations of the distance between the microphones L (can be set by consumers), the required angle α and the frequency band of the formants can be applied to a wide range of sound receivers.

It is clear that only in the presence of four microphones is it possible to create a receiver of human speech with a sharp focus. If the sound source has one, two or more formants (critical frequency bands), the number of microphones, amplifiers and filters also changes. Figure 6 shows a sharply directed sound receiver in accordance with the proposed method, in which the end microphones 9 with a distance Lmax are designed to register the low-frequency sound components of the formant Ф ₁ . Mid- and high-frequency distances Lcp and Lmin are possible when microphones are turned on inside the distance Lmax. The formation of pairs of microphones with the inclusion of a microphone-summing amplifier of filters 15'-15 'into the middle range of the formant Ф ₂ and 15''-15''to the high-frequency range is shown in Figs. 3 and 6. Thresholds at the amplifier inputs are not shown, so as their introduction is not a requirement.

As shown in figure 3, the authors propose operational amplifiers, formant filters, a power source, possible switches and adjustments to be placed inside a hollow special handle 11, which is equipped with a special structural element for mounting the receiver in a stationary position on any support. It should also be noted that according to the authors, the main components and materials for the implementation of the proposed method are publicly available and do not require special developments, not counting the design. So, you can recommend condenser type electret microphones as sound receivers. Electric filters are the most dimensional elements, but their range, as well as summing operational amplifiers on the electronic sites of developers, manufacturers is also extensive.

Thus, the authors believe that the proposed method of sharply directional reception of sound waves perceived by the human hearing, allows to reduce the geometric dimensions of the receiver and improve the directivity.

Literature

1. Aldoshin I.A., Vologdin E.I., Efimov A.P. and others. "Electroacoustics and sound broadcasting." - M .: Hot line - Telecom, 2007.865 s.

2. Vakhitov Sh.Ya., Kovalgin Yu.A., Fadeev A.A., Shcheviev Yu.P. "Acoustics". - M .: Hot line - Telecom, 2009 .-- 660 p.

3. Kuznetsov A.A. "Acoustics of musical instruments." - M.: Art, 1989.

4. Sapozhkov M.A. "A speech signal in cybernetics and communication." - M .: Svyazizdat, 1963 .-- 452 p.

5. Mikhailov V.G., Zlatoustova L.V. "Measurement of speech parameters." - M .: Radio and communications, 1987. - 168 p.

6. Sapozhkov M.A. "Electroacoustics". - M.: Communication, 1978.- 272 p.

Claims (1)

A method of highly directional reception of sound waves, in which the reception is carried out by receivers located on a rigid linear basis, characterized in that at least four microphones are used as the receiving sound, the zero direction of the receiver is determined by the normal position of the axis of the sound source — the middle of the base, and the outputs of the pair microphones are connected to two inputs of the operational amplifier through electric filters at frequencies that correspond to the critical frequency bands of a person’s hearing, and the number of opera at least three radio amplifiers, and the microphone outputs with a maximum distance between them L _{max are} connected to the operational amplifier through low-pass filters, pairs of microphones located at a distance of L _cf and L _min , respectively, are connected through medium and high-frequency filters, the output signals of the mentioned operational summing amplifiers are used to form a single output signal, and the values of these distances are set from the expression

Where

- phase shift at the outputs of a pair of microphones when the front of the sound wave deviates from the zero position by an angle ±

expressed as fractions of the sound wavelength

, C _sv and f _sv are the speed of sound and the frequency of the component of the sound signal, respectively.

Images